Demographics

Column

Co-morbidity and Prior Cardiovascular Event Frequencies

Column

HF Admissions by Year

Summary Statistics

Cohort Characteristics
Readmitted Not Readmitted
ami hf stroke tia other noreadm
Insurance Coverage
Commercial 683 (43%) 937 (38%) 471 (39%) 323 (41%) 798 (38%) 706 (38%)
Medicare 742 (47%) 1,259 (51%) 616 (51%) 388 (49%) 1,070 (51%) 996 (53%)
Other 158 (10%) 252 (10%) 130 (11%) 78 (10%) 214 (10%) 179 (10%)
Demographics
Avg Age (SD) 59 (16) 59 (16) 59 (16) 59 (16) 59 (16) 59 (16)
Sex (% Male) 61% 60% 60% 60% 59% 58%
N 1,583 2,448 1,217 789 2,082 1,881

Model Results

Column

30-day All-Cause Readmission Model

Column

ROC Curve

Model Coefficients

About

The data set contains simulated hospital admission records of 10,000 heart failure (HF) patients. Each patient has a single record corresponding to the patient’s first HF admission (i.e. the index admission).

Abbreviation Description
Patient_ID Patient ID (1 through 10,000)
bdat Date of birth
admdat Index admission (I.e. first HF admission) date
sex Sex
insur Primary insurance type
readmreason Reason for readmission: HF, AMI, Stroke, TIA or Other
readmdat The date of first readmission after the index admission
deathdat Date of death

Study endpoints include all-cause death, all-cause readmission and HF readmission. The outcomes are not explicitly defined: NAs in deathdat and readmdat columns signify that the event did not occur until the end of the study (31/12/2015). Each event can be defined within a given time interval (e.g. 30, 60, 90, 180 days, 1 year from the index admission, etc.) and analyzed using, for example, logistic regression. Since both, the date of the index admission and the date of an event are known, survival analysis can also be done. Think of using birth dates as well, e.g. age at the index admission as a covariate in the model, age cohorts, etc.). For the all-cause readmission, the distribution of reasons for the readmission can also be studied.

Comorbidities defined at the time of index admission. These include chronic conditions (hypertension, COPD, diabetes, etc.) as well as prior cardiovascular events (AMI, stroke, TIA, etc.). You can define your own inclusion/exclusion criteria, e.g. cancer.

Abbreviation Description
ami Acute Myocardial Infarction
stroke Stroke
tia Transient Ischemic Attack
hyp Hypertension
asc Atherosclerosis
cmyo Cardiomyopathy
anem Anemia
lipid Disorders of Lipid Metabolism
diab Diabetes
obes Obesity
cld Chronic Liver Disease and Cirrhosis
kd Chronic Kidney Disease/Acute Kidney Failure
copd Chronic Obstructive Pulmonary Disease
sleep Sleep Apnea
thyr Thyroiditis
cancer Cancer